Prime Intellect is hiring a Member of Technical Staff - Inference

Prime Intellect is hiring a remote engineer for its inference stack. The company builds the open superintelligence stack — compute, environments, evaluations, sandboxes, high-performance training, and deployment in one platform — and trained INTELLECT-2, a 32B reasoning model, with fully asynchronous RL across GPUs on three continents.

What you'll do

• Build and optimize the inference layer serving open frontier models at scale
• Work across the boundary between training and deployment, where post-training output becomes a production endpoint
• Squeeze throughput, latency, and cost out of distributed, heterogeneous GPU fleets

Requirements

• Strong systems engineering with modern inference engines and GPU performance work
• Comfort with distributed infrastructure that spans providers and regions
• Interest in open models and decentralized compute

Remote. Full details and application at the link.